NVIDIA Run:ai Is Now Open Source as Part of NVIDIA DSX AI Factory Platform | NVIDIA Technical Blog

This post announces the open-sourcing of the NVIDIA Run:ai platform, which includes KAI Scheduler, GPU Fractioning Operator, KAI Resource Management, and Workload Map. KAI Scheduler is highlighted as the core Kubernetes-native scheduler for GPU workloads, contributing to the CNCF Sandbox. Key features of KAI Scheduler include gang scheduling, fair-share and quota allocation, topology awareness, GPU sharing (MIG, MPS, time sharing, HaMi, CUDA 13.4 MPS fractioning), and Dynamic Resource Allocation (DRA) support. The release aims to accelerate innovation and bring AI factory solutions to market faster by providing an open foundation for AI workload and GPU orchestration.